So when used to describe servers, an example would be if the two web sites I showed were actually hosted on the same physical hardware. They wouldn’t actually be two servers, but they would “virtually” appear to be running whole autonomously of one another. This is done either by hosting multiple web sites in IIS through the host header feature and/or multiple IP features or by using a virtual server product, which actually hosts a whole other operating system within the server’s primary OS (thereby giving you multiple full-blown server OS instances running on one physical hardware box – each server instance is therefore “virtual”).
When used to describe a directory, it essentially means a URL that doesn’t physically exist under the root path but rather is mapped that way. Let’s take the document’s example. If you had your web root path set to C:inetpubwwwroot and you had a bunch of files at D:files, you could create the virtual directory in IIS named “documents” that points to D:files. So while C:inetpubwwwrootdocuments doesn’t actually physically exist, you could browse to http://www.ivrtechgroup.com/products-services